A method for water level detection based on image inverse perspective transformation

Abstract

The application discloses a kind of based on image inverse perspective transformation and realizes water level detection method, belong to the technical field of computer vision application, the method comprises the following steps: S1: water level gauge detection obtains water level gauge overall target area and water level gauge scale target area;Water surface segmentation result is also obtained to water surface segmentation simultaneously;S2: according to water level gauge scale detection frame and water level gauge detection frame, inclination is judged, and pixel distance is calculated;S3: pixel distance between water level gauge detection frame top and water level line is used, pixel is combined with the real distance represented to calculate water level gauge real height.The present application makes full use of the application in the field of image of computer vision, carries out water level gauge detection and water surface segmentation using deep learning model, realizes water level gauge inclination correction in image by inverse perspective transformation, obtains water level line position by water level gauge and water surface segmentation result intersection, to obtain accurate water level value.

Images